What’s the future of AI and music? Things may be starting to become clear

At this year's All That Matters conference in Singapore, AI and music was a major topic. How will things eventually shake out? Here are some predictions.

Every September, leading up to the Formula One race, Singapore hosts a conference called All That Matters. Thinkers, creators, and entrepreneurs from all over Asia gather to talk about the latest developments in sports, tech, marketing, gaming and music. I try to go every year, because in many aspects, Asia is already living in the future. (Actually, that’s incorrect. They’re living in the present; it’s parts of the West that are clinging to the past. But I digress.)

Artificial intelligence was front and centre for many of the sessions, especially those dealing with music. There was much discussion about how AI will factor into things like music creation, music distribution and copyright. I took notes.

The internet we’ve been used to is once again changing rapidly. The browser wars are back, too, thanks to searches turbocharged by AI. I’ve been trying out Comet, a brand-new browser from the people behind the AI program Perplexity. It still has some rough edges, but I can see myself liking it better and using it more than Google. Then again, Google has its own AI in the form of Gemini, which has been going up against Microsoft’s Co-Pilot.

The last quarter century has been about changes in consumption: CDs to MP3s to piracy to streaming, with a detour back into vinyl. The next decade will see major shifts in creation. Ignore and disparage that all you want, but it’s going to happen.

When it comes to artificial music creation, the big players are Suno and Anthropic, two generative AI programs that keep getting better when it comes to making music based on user prompts. Although both are deep in litigation with rightsholders and labels, expect licensing deals to be hammered out with the major labels sooner rather than later. This means more human-created music will be used to feed the models behind AI programs. By this time next year, these programs will be able to analyze trillions of data points when responding to prompts. More data points means (theoretically) more realistic and — this is key — emotional music. Predictions are that this will set off a new era for the recorded music industry.

One way that will happen is a simplification of user interfaces. One of the biggest problems with any computer program is learning how to use it. Today, the thinking is the less user interface (UI), the better. We’re approaching a situation where the programs will say, “Just tell me what to do and I’ll do it by pulling all the tools that exist in the cloud.” Already, we have programs that will return produced music if you just hum an idea into it.

The bulls outweighed the bears on this one. Here’s how AI might benefit musicians very soon.

Whenever any kind of promising technology comes along, there has always been a segment of humanity that will drag it down to its lowest common denominator. (Hello, social media!) There will also be knock-on effects. For example, if generative AI programs blow up the way they’re expected to, what happens to real-life musical instruments and the people who make them? What happens to traditional recording studios and the people who staff and equip them? Will people of the future bother to learn to play a traditional instrument by putting in 10,000 hours of practice?